Optimization of roadside management based on artificial intelligence
Industrial engineering, system engineering, computer science
The relationship between road infrastructure and economic development is widely studied, as 25 million kilometres of new roads will be built by 2050. The vegetated strips of land located near roads that separate them from the surrounding landscape, called roadside or road verges, are affected by this growth in road infrastructure. Recent studies have shown that the roadside plays an important role in mitigating road impact on the territory and its ecosystems. A sustainable management approach of these green spaces can provide different ecosystem services that can improve the citizen’s life conditions. Indeed, the maintenance of roadsides integrates several issues with various aspects: (i) economic (e.g. continuous investment in the maintenance equipment, the attractiveness of the territories linked to the quality of the maintained landscape), (ii) technological (e.g. the biomass valorisation, the reduction of the carbon footprint), (iii) social (e.g. road safety, flood and fire prevention) and (iv) environmental (e.g. the preservation of the biodiversity, the water improvement, the air and soil quality, etc.).
Optimal road mowing is a significant challenge for territories, which requires a large budget and detailed annual maintenance in terms of staff and mowing devices; making informed decisions regarding the staff requires a minimum distance that needs to be delivered in a limited amount of time to represent crucial factors that need a data-driven decision-making modelling approach. This thesis subject addresses the problem of planning optimal trajectories of roadside maintenance machines (mainly mowing machines) in order to minimize economic and environmental costs and impacts on the territory. In this project, territories will be able to serve as territories of experimentation.
The project comprises of three major phases:
1) Investigating and collecting all data related to the road side maintenance from several regions in France.
2) Building a multi-objective optimisation algorithm that will find the minimal route that needs to be undertaken in order to service all road types in a dedicated area, by their own dedicated technical centre.
3) Construct several scenarios of optimisation approaches based on time and human constraints that would lead to a reduction in fuel consumption and allocated time throughout the year.